Improvements and fixes for the LUKE integration #13
add model_size argument to the conversion script
9df2c09b
add support binary cross entropy loss in all head models
fae11db0
rename [ENT] to <ent>
ad68db00
improve LukeTokenizer (work in progress)
9d97a26f
bug fixes
9d34d114
update the conversion script to use the tokenizer's encode function
82351a25
remove tokenization_luke_fast.py
a57f87d3
add docstring
1926a202
fix issues in tokenizer
94cdf00d
fix model names and urls
1c9d7a33
NielsRogge
merged
015b2099
into adding_luke_v2 4 years ago
Assignees
No one assigned
Login to write a write a comment.
Login via GitHub